Meni8permacefe.~] FLORA TNDICA. 1B9 The stamens are normally free and definite, one being placed opposite each petal, so that they form two'verticils. In Zimacia triandra they are reduced to three; in another species of the same genus their numBer is nine; and in Menupermum and Calycocarpum fihey are indefinite. In Pycnarrhena, Chamanthera, and Abuta (of Poppig) the filaments are partially mouadelphons j and in Parabana, Aspidocarya, and the whole of the tribe Cissaai$elidea> they are combined into a central column, bearing on its apex a flat peltate disc, antheriferoas round the margin. In Ana- vnirta, the structure is still more complex, the anthers being united into a globose mass. The ovaries arc sometimes seated directly on the torns, but not un frequently they are supported by a. distinct gynophore, which becomes very conspicuous as the fruit advances to matuijjy. Tbeir number is usually three; but in the tribe dssampeti- dets they are nlways solitary, and in Cocculzts ovalifolius and Coscimum there are generally six. In TUiacora they arc indefinite in number. * i The ovary of Memspermacea is generally oval or oblong, straight on the ventral suture, and rounded on the back, with a terminal style. The ovules are solitary and peltate, and ^inserted at or below ihe middle of the ventral suture, with the micro- pyle invariably superior and the chalaza at the broad end of the ovule, which is nearest the base of the ovary. In Aspidocarya, and an undetermined species nearly allied to it, in" which the seed is pendulous and anatropons, the ovule is probably at- tached near the apex of the ovary; but nevertheless the micropyle and foramen have the same position as iu the rest of the Order. During the ripening of the fruit great changes take place in the structure of the avary. The dor$u$i -grows more rapidly than the' ventral part, so that the style or its cicatrix, which ,is terminal in the ovary, is in the ripe fruit more or less lateral, and in, a large part of the Order is situated close to the base of the carpel. \Vhile this irtregular development of the parietes of the ovary is proceeding, the inner wall gradually hardens into a more or less woody putamen, sometimes very thick and almost bony, at other -times thin and brittle, and variously tuberculated. At the same time the podosperm lengthens as the hilum of the seed is carried by the in- creasing curvature of the walls of the ovary further and further from the base of the fruit; while the putamen, which thus becomes as it were doubled upon itself, in- vests it with & bony sheath, which takes a great diversity of form in different parts of the Order. Mr. Griffith* lias thrown out a conjecture that the woody or bony portion of the fruit is not putamen, but testa. This view receives some support from the fact that only one very delicate coat can be detected an the seed, and from the peculiar mode in which the bony coat adapts itself to the shape of the seed j but it is not borne out by a study of the development of the ovule, which we have been able to trace so satisfactorily as to ascertain beyond a doubt that this coat belongs to the ovary, and not to the ovule. The form of the embryo is very different in different tribes of the Order. Except in Aspidocarya it is always more or less curved; and in the greater part of'the Order, where the 'style-scar is situated near the base of the fruit, the radicle, which always points towards it, is brought almost into contact with the base of the fruit aud the chalazal extremity of the seed. In the division Heteroclinea the cotyledons are foliaceous and very thin, and (usually laterally) divaricated, so as to occupy dis- tinct cavities in the albumen.. The seed is therefore broad, and, but for the peculiar mode of growth of the putamen, would be quite flat, as it is in the genus Aspido- carya. This, however, causes it to assume a globular shape; but it IB hollow within, and moulds itself on an internal process of the putamen, which Mr. Miers has called condyle.^ In the remainder of the Order the narrow, strap-shaped or hcmi- * Itinerary Notes, p. 165. f "^e have not adopted this term, partly because it does not represent an organ or structure analogous to that so called in osteology, and partly because we hesitate to apply specific terms to modifications of siyacture which arę confined to small 7.